Comparison of Auditory Brainstem Responses Recorded from Three Different Clinical Systems

Abstract

Background:

The auditory brainstem response (ABR) test is used to estimate hearing thresholds and for neurodiagnostic purposes. ABRs can be recorded for traditional stimuli such as broadband clicks and frequency-specific tone bursts (0.5, 1, 2, and 4 kHz). Recently, chirp stimuli (broadband and narrowband chirps [0.5, 1, 2, and 4 kHz]) have been recommended to be used to record ABRs.

Purpose:

There is a lack of comparative studies of ABR latency, amplitude, and thresholds obtained from these systems for both traditional and chirp stimuli.

Research Design:

Repeated measure design.

Study Sample:

Fifteen adults whose hearing thresholds were ≤25 dB HL participated in this study.

Data Collection and Analysis:

The Interacoustics Eclipse-25 (EP-25), the Intelligent Hearing System (IHS), and the Vivosonic Integrity V500 systems were used. Peak-to-peak voltage was examined for traditional and chirp stimuli across all three systems. Behavioral and ABR thresholds were obtained for all stimuli. Also, peak V latencies and peak-to-peak amplitudes for all stimuli recorded at 80 dBnHL were examined.

Results:

Peak-to-peak voltage differed between systems, especially for chirp stimuli. The EP-25 and the IHS systems showed significantly lower behavioral thresholds for some chirp stimuli compared with traditional, comparable stimuli, whereas the Integrity V500 system showed significantly lower behavioral thresholds for broadband click and 4-kHz tone burst stimuli compared with comparable chirp stimuli. Significant differences in behavioral thresholds between systems were observed. The EP-25 (broadband chirp, 1-, 2-, and 4-kHz narrowband chirp) and the IHS (2-kHz narrowband chirp) showed significantly lower ABR thresholds for comparable stimuli, whereas the Integrity V500 system showed lower ABR thresholds for broadband click and 0.5-kHz tone burst stimuli compared with comparable chirp stimuli. ABR thresholds for 2-kHz tone burst and 0.5- and 2-kHz narrowband chirp stimuli were significantly different between systems. The peak V latencies of the IHS system were significantly longer for most of the stimuli compared with other systems. The Integrity V500 system showed significantly lower peak-to-peak amplitudes of peak V for most of the stimuli when compared with other systems.

Conclusions:

The ABR latencies, amplitudes, and thresholds may differ between ABR recording systems, and clinicians should be aware of these differences to avoid misdiagnosis.

Clinical Relevance Statement:

The differences in ABR latencies, amplitudes, and thresholds between systems may lead to inconsistencies in diagnostic accuracy. Audiologists should use system-specific normative data.

INTRODUCTION

The auditory brainstem response (ABR) test is used to estimate hearing thresholds and for neurodiagnostic purposes (Joint Committee on Infant Hearing, 2019). The ABR can be recorded by presenting short-duration acoustic stimuli (traditional stimuli: broadband 100-microsecond click, frequency-specific tone bursts; chirp stimuli: broadband chirp and narrowband chirp stimuli) (Hall, 2007). The ABR consists of five major positive peaks (I, II, III, IV, and V), and the latency and amplitudes of these peaks are examined. The latency reflects how long it takes for stimuli to trigger neural activity in the auditory brainstem pathway. The amplitude represents the magnitude of neural activity in response to the stimulus. ABR peak latencies are shorter at higher intensities, and amplitudes are larger; whereas at lower intensities, latencies are longer and amplitudes are reduced. Only peak V is reliably present at lower intensity levels and exhibits greater resistance to background electroencephalography (EEG) noise. The peak V is a robust and repeatable component of the ABR, commonly used either in isolation or in conjunction with other waves for ABR analysis (Hall, 2007).

Traditionally, ABR is recorded using broadband click (100 microseconds) and/or frequency-specific tone burst stimuli (0.5, 1, 2, and 4 kHz) for neurodiagnostic and/or threshold estimation. The click stimulus is commonly used to record ABRs for neurodiagnostic purposes (Allen & Allan, 2014; Ankmnal-Veeranna et al., 2019; Hall, 2007; Hood, 1998; Veeranna et al., 2021, 2022). For neurodiagnostic purposes, click-evoked ABRs are usually recorded for different stimulus rates (slow vs. fast). If the peak V latency is not within reference limits or the latency shows an abnormal shift at faster stimulation rates, it is considered clinically abnormal (Allen & Allan, 2014; Ankmnal-Veeranna et al., 2019; Hall, 2007; Hood, 1998; Veeranna et al., 2021, 2022). Either the peak-to-peak amplitude of peak V or the amplitude ratio (peak-to-peak amplitude of peak V/peak-to-peak amplitude of peak I) is examined (Allen & Allan, 2014; Ankmnal-Veeranna et al., 2019; Hall, 2007; Hood, 1998; Veeranna et al., 2021, 2022). The peak V (peak-to-peak) amplitude is usually larger than or equal to the peak I (peak-to-peak) amplitude. The peak-to-peak amplitude is variable among participants. Hence for neurodiagnostic purposes, V/I amplitude ratio is examined (Hall, 2007). It is clinically considered abnormal if the V/I amplitude ratio is less than 0.50 (Hall, 2007).

For threshold estimation, frequency-specific ABRs are recorded. In frequency-specific ABRs, the peak V latency is longer for lower frequencies than for high-frequency tone burst stimuli (Dau, 2003; Gorga et al., 1988). The peak V latency and amplitude are often compared with stimulus intensity when analyzing the ABR. However, although the peak V latency is not usually measured in frequency-specific ABRs, it can still provide an idea of where the peak occurs, if measured.

Traditional stimuli (click and frequency-specific tone bursts) have been audiologists’ main choice for threshold estimation for years. However, traditional stimuli do not elicit robust peak V at lower intensity levels (Cobb & Stuart, 2016) because click stimuli, due to transducer characteristics, excite maximally the basal end of the cochlea and not the entirety. Hence, to address this limitation, researchers have designed a new stimulus called a chirp (Elberling & Don, 2008). Chirp stimuli produce a rapid sweep from low to high frequencies, designed to activate simultaneously a wide range of cochlea, from base to apex. This is achieved by compensating for frequency-dependent travel-time differences, ensuring low frequencies enter first, followed by mid and high frequencies. As a result, auditory nerve fibers for all frequencies are activated at the same time, avoiding phase cancellations across frequencies and enhancing neural synchrony, increasing the amplitude of the composite neural response. Different types of chirp stimuli are designed based on models of the cochlear delay (Dau et al., 2000; de Boer, 1980; Fobel & Dau, 2004). A broadband chirp stimulates nearly the entire basilar membrane, but a more frequency-specific (0.5, 1, 2, and 4 kHz) narrowband chirp is designed so that it excites a frequency-specific portion of the basilar membrane. A narrowband chirp stimulus is created by using octave bands that limit the frequency output to be centered around a much smaller range of frequencies. Broadband and narrowband chirp stimuli have recently become the recommended choice of stimuli for threshold estimation because they elicit a larger amplitude, especially at threshold levels, and reduce the overall test time (Ferm & Lightfoot, 2015; Sininger et al., 2018). However, when compared with traditional frequency-specific tone bursts, narrowband chirp stimuli are one octave wide, activating more regions on the basilar membrane and eliciting larger amplitude, but affecting frequency specificity (Adjekum & Stapells, 2023; Wegner & Dau, 2002). In addition, the longer duration of chirp stimuli can cause ABR latencies to differ from those evoked by traditional click and tone burst stimuli.

A variety of US Food and Drug Administration (FDA)-approved ABR recording systems are available. These ABR systems use different types of chirp stimuli to evoke ABRs (refer to Table 1). The Eclipse-25 (EP-25; Interacoustics, Middelfart, Denmark) uses the Claus Elberling Chirp® (CE-Chirp®). The CE-Chirp is designed based on derived-band ABR latencies obtained from a large group of adults whose hearing thresholds were within 20 dB HL (Elberling & Don, 2008). The broadband CE-Chirp has a flat-amplitude spectrum from 0.35 to 11.3 kHz. From the broadband chirp, four octave-band filtered versions of the CE-Chirp were constructed with center frequencies of 0.5, 1, 2, and 4 kHz (Cebulla & Elberling, 2010; Elberling & Don, 2008). Compared with traditional stimuli, CE-Chirp stimuli are longer in duration. The duration of the CE-Chirp stimulus is level specific (LS), that is, shorter at a higher intensity (>60 dB nHL) and longer at lower intensity levels (Elberling & Don, 2010). The EP-25 system time-adjusts the CE-Chirp ABR for broadband CE-Chirp and narrowband CE-Chirp to match the latencies of traditional stimuli.

Table 1.

ABR Recording Parameters

	EP-25	IHS	Vivosonic
Type of averaging	Bayesian weighted average	Conventional averaging	Kalman weighted average
Inserts	IP30	ER3C	ER3C
Rate	39.1/s	39.1/s	39.1/s
Filter	0.033–3 kHz (6 dB/octave)	0.03–3 kHz (6 dB/octave)	0.03–3 kHz (high-pass: 12 dB/octave; low-pass: 24 dB/octave)
Traditional stimulus	BB-Click, 0.5, 1, 2, and 4 kHz	BB-Click, 0.5, 1, 2, and 4 kHz	BB-Click, 0.5, 1, 2, and 4 kHz
Tone burst duration (Blackman gated, cycles)	2-0-2	2-0-2	2-0-2
Chirp stimulus	BB-(LS)-CE-Chirp, LS 0.5-, 1-, 2-, and 4-kHz NBC	BB-iChirp, 0.5-, 1-, 2-, and 4-kHz NBC	BB-VF Chirp, 0.5-, 1-, 2-, and 4-kHz NBC
Chirp duration	N/Aa	BB: 3.95 ms 0.5 kHz: 5.00 ms 1 kHz: 5.00 ms 2 kHz: 3.00 ms 4 kHz: 2.00 ms	BB: 12 ms 0.5 kHz: 10.93 ms 1 kHz: 8.64 ms 2 kHz: 5.02 ms 4 kHz: 3.46 ms
Polarity	Rarefaction for BB-Click and Chirp. Alternating for TBs and NBC.	Rarefaction for BB-Click and Chirp. Alternating for TBs and NBC.	Rarefaction for BB-Click and Chirp. Alternating for TBs and NBC.
Intensity	80, 60, 40, and 20 dB nHL and threshold search (5–10 dB)	80, 60, 40, and 20 dB nHL and threshold search (5–10 dB)	80, 60, 40, and 20 dB nHL and threshold search (5–10 dB)

Note: Chirp stimulus duration information for the IHS and Vivosonic Integrity V500 systems was found in the technical bulletin document. BB = broadband; N/A = not available; NBC = narrowband Chirp.

^aExact stimulus duration information for the Interacoustics EP-25 BB, 0.5, 1, 2, and 4 kHz could not be found in user manual.

Studies have reported that ABR thresholds are lower for CE-Chirp stimuli (broadband and narrowband CE-Chirp) when compared with traditional stimuli (broadband click and frequency-specific tone bursts) (Bargena, 2015; Ceylan & Kaya, 2023; Cobb & Stuart, 2016; Ferm et al., 2013; Ferm & Lightfoot, 2015; Michel & Jørgensen, 2017; Rodrigues et al., 2013). The time taken to estimate hearing thresholds is less for CE-Chirp stimuli when compared with traditional test stimuli (Bargena, 2015; Ferm et al., 2013; Ferm & Lightfoot, 2015). Studies have also reported good test-retest reliability for CE-Chirp stimuli (Cobb & Stuart, 2014; Jamal et al., 2021). The estimated HL (eHL) is the behavioral audiometric threshold (in dB HL) that is estimated from the ABR thresholds (in dB nHL). The dB eHL or the correction values available for narrowband CE-Chirp from the EP-25 system are roughly 5 dB less than that of dB eHL correction values for tone bursts (Ferm et al., 2013; Sininger et al., 2020). Sininger et al. (2020) reported dB eHL correction values for narrowband CE-Chirp recorded from the EP-25 system for different degrees of hearing loss.

Each of the Intelligent Hearing Systems (IHS, Miami, FL) devices that record ABR uses broadband and narrowband chirp stimuli (iChirp). iChirp stimuli are generated based on the de Boer linear model (de Boer, 1980). The iChirp stimulus is designed such that it elicits larger ABRs compared with ABRs evoked for traditional stimuli. The broadband iChirp has a similar spectrum to that of click stimulus, and octave band versions of narrowband-iChirp are centered around 0.5, 1, 2, and 4 kHz. The duration of broadband-iChirp is similar across all intensity levels. The IHS does not time-adjust the display of ABRs for broadband-iChirp and narrowband-iChirps. Therefore, the ABR latencies generated by these stimuli would be longer relative to those of traditional stimuli. Narrowband iChirp stimuli elicit lower thresholds when compared with traditional stimuli (Pinto et al., 2022).

The Vivosonic Integrity-500 (Toronto, ON, Canada) uses broadband and narrowband chirp stimuli (VF Chirp). The VF Chirp stimuli meet International Electrotechnical Commission (IEC) 60645-3 (2020) specifications. The broadband VF Chirp has a similar spectrum to that of a click stimulus. Octave band versions of narrowband VF Chirp are centered around 0.5, 1, 2, and 4 kHz. The duration of the broadband VF Chirp stimulus is similar across all intensity levels. The Integrity V-500 system does time-adjust the ABR to match the latencies of traditional stimuli. Research on similarities and differences in ABRs recorded for traditional and VF Chirp stimuli in individuals whose hearing thresholds are ≤25 and ≥25 dB HL is scarce.

The earlier-mentioned FDA-approved clinical ABR recording systems are readily available. These systems use different types of chirp stimuli, algorithms to record and analyze evoked responses, and response detection methods. Consistent ABR latencies and amplitudes between systems are critical for auditory neurodiagnostic purposes and for hearing threshold estimation. ABR peak latencies serve as diagnostic indicators and reflect the overall integrity of the auditory brainstem. If there are differences in latencies and amplitude between systems, it will compromise the reliability and reproducibility of test results. Hence similarities in ABR latencies between systems are crucial for neurodiagnostic purposes to avoid misdiagnosis. Authors recommend using published normative data for ABR absolute latencies (Hall, 2007; Hood, 1998; Sininger, 1992; Stapells, 2011), provided the tester’s ABR system matches the one used in the study. The ABR system’s sampling points should also be similar to ensure accuracy because higher sampling points may yield more precise latency measurements than lower ones. However, currently, it is not clear whether ABR latencies differ between systems.

Estimated hearing thresholds should be similar if different systems are used, as long as the test protocol, stimulus, and test environment are identical. It is also recommended that any new technology or stimulus used for objective estimation of frequency-specific hearing thresholds in infants should be validated rigorously for its accuracy (Joint Committee on Infant Hearing, 2019). However, currently, studies are scarce comparing ABR thresholds obtained from different FDA-approved systems. A recent literature review reveals that only two recent publications (Bagatto et al., 2024; Leusin Mattiazzi et al., 2024) address this scarcity at the time of writing this manuscript. Leusin Mattiazzi et al. (2024) compared narrowband chirp-evoked ABRs (0.5, 1, 2, and 4 kHz) recorded from the IHS and the EP-25 systems. ABRs were recorded from 20 children (6 months to 12 years old) whose hearing thresholds were ≥25 dB HL. The ABR thresholds for the narrowband CE-Chirp and iChirp stimuli were comparable among the participants. Bagatto et al. (2024) compared the ABR thresholds obtained for frequency-specific tone bursts (0.5, 1, 2, and 4 kHz) from the Vivosonic Integrity V500 with the Biologic Navigator Pro systems. ABRs were recorded from 37 adult and 105 infant ears with ≤25 and ≥25 dB HL hearing thresholds. The authors reported no significant differences in ABR thresholds between systems. The findings of these two studies suggest that ABR thresholds obtained from two different systems are comparable. The EP-25, IHS, and Integrity V500 systems employ different types of chirp stimuli to record ABRs. However, we are not aware of any studies that have compared the responses of these three systems with chirp stimuli to determine whether they yield similar results. In this study, we compared the outcomes of three different ABR recording systems (EP-25, IHS, and Integrity-V500). Behavioral and ABR thresholds were obtained for traditional stimuli (broadband click, 0.5-, 1-, 2-, and 4-kHz tone bursts) and chirp stimuli (broadband and 0.5-, 1-, 2-, and 4-kHz narrowband chirps) using these three systems from adults whose hearing thresholds were ≤25 dB HL. Also, ABR latencies for traditional and chirp stimuli recorded at 80 dB nHL were compared between the three FDA-approved ABR systems.

METHODS

Participants

Fifteen adult participants (mean age: 25.49 years, age range: 22.21–37.85 years) were enrolled in this study. Participants did not report any significant history of noise exposure, middle ear infection, neurological dysfunction, or a family history of hearing loss. The pure tone audiometry (Two-channel GSI AudioStar-Pro) test was conducted to obtain air and bone conduction thresholds for each participant. A Hughston-Westlake (Carhart & Jerger, 1959) procedure was used to obtain hearing thresholds. All participants' thresholds at octave frequencies (0.25–8 kHz) were ≤25 dB HL in both ears. The tympanometry test (Interacoustics, Titan) was conducted, and all participants showed the presence of a type-A tympanogram in both ears.

ABR Recording

In this study, EP-25, IHS, and the Integrity V500 ABR recording systems were used. Both EP-25 and Vivosonic Integrity V500 systems are two-channel systems, whereas the IHS is a four-channel system. For this study purpose, only two channels were used. Table 1 shows the ABR recording parameters used in this study. The EP-25 system was calibrated by a technician. The IHS and the Integrity V500 systems were calibrated by the manufacturer. These three systems differ in the technology used to record ABRs; however, as shown in Table 1, attempts were made to closely match the test protocol used with each system.

Behavioral Thresholds for Test Stimuli

Before recording ABRs from the participant, behavioral thresholds for the test stimuli (39.1/second [s]) were obtained from the test ear. Participants were instructed to respond by raising their hands whenever they heard a sound. For each participant, a minimum of 200 stimuli were presented, but the presentation was stopped if the participant raised their hand to indicate that they heard the sound. In the absence of a response from the participant, up to 500 stimuli were delivered before increasing the intensity by 5 dB. The initial presentation level was 20 dB nHL. A 10 dB down and 5 dB up procedure was used to estimate the lowest level at which they could recognize the sound. Test stimuli were chosen randomly.

ABR Thresholds for Test Stimuli

ABRs were recorded using a vertical montage. Inverting electrodes were placed on both earlobes, the noninverting electrode was placed on the upper forehead, and the ground was placed on the forehead. The impedance of these electrodes was <5 kΩ. The ABR recording parameters reported in Table 1 were used. Stimulus intensity was 80, 60, 40, 20, and a threshold search (5–10 dB steps). The ABR recording was stopped if one or more of the following criteria were met: the peak V was visible and followed an expected latency intensity function, the Fmp (=3.25) (EP-25 system), the signal-to-noise ratio (=1 or more) (IHS and the Integrity V500), a repeatable peak V in both buffers, and the peak V was replicable at threshold levels. The ABRs were recorded below the threshold level and confirmed the absence of peak V at these subthreshold levels. Latencies and peak-to-peak amplitudes of peak V for ABRs recorded at 80 dB nHL were analyzed. All ABR markings were agreed on by all of the authors. Behavioral thresholds for the test stimuli and ABR thresholds were obtained from only one ear, and the test ear was randomly chosen.

The University of Southern Mississippi Institutional Review Board has approved the study methods. All participants signed the informed consent form. All ABR recordings were completed in a sound-treated room. The ambient noise levels measured in the test room were well below the maximum permissible levels recommended by the American National Standards Institute (ANSI) standards (with ears covered for insert earphones, ANSI S3.1-1999 [R2008]). All participants were encouraged to schedule a time when they would be most likely to effectively participate in ABR testing. Each scheduling period was set up to complete an assessment for one system per session. Each session lasted a maximum of 2 hours (h; ∼6 h of testing for each participant).

Statistical Comparison

The Shapiro-Wilk test was used to determine whether the data were normally distributed. If the data were normally distributed, a paired t test (parametric) was used; otherwise, the Wilcoxon test (nonparametric) was used to examine differences in behavioral thresholds for traditional and chirp stimuli. Levene’s test of equality of variances was conducted. If Levene’s test was not significant, a parametric test was used (three or more means comparison: one-way analysis of variances, post hoc comparison: t test); otherwise, a nonparametric test (three or more means comparison: Kruskal-Wallis; post hoc comparison: Mann-Whitney U test) was used. For all the analyses, the significance value of p < 0.05 was used. For multiple comparisons, the p value was adjusted. All analyses were conducted in JASP (0.19.2) (2024).

RESULTS

Figure 1 shows the mean peSPL (peak equivalent sound pressure level) values corresponding to 0 dB nHL for different stimuli from different ABR recording systems. These data are stored in the manufacturer’s software (IHS and Integrity V500 systems) and are available in their user manual (EP-25 system). Within each manufacturer, decibels peSPL (dBpeSPL) values for traditional stimuli are different when compared with chirp stimuli. For example, the EP-25 system’s dBpeSPL values are 35 and 31 for broadband click and chirp stimuli, respectively. Mean peSPL values are comparable for traditional stimuli, but not for chirp stimuli between systems.

Figure 1.

Mean peSPL to nHL correction values of test stimuli for different ABR recording systems. These data are available in the manufacturer’s software (IHS and Vivosonic systems) or the user manual (EP-25 system).

The sound output of the three systems was examined. The microphone (1/2 inch) was placed in a 2-cubic centimeter coupler (AEC-202) and attached to a sound level meter (731 C; Larson and Davis). The microphone was calibrated by attaching the acoustic calibrator (SC-942; Lutron) to the microphone and presenting a 94 dB SPL pure tone at 1 kHz. The sound level meter reading was checked to ensure it matched the calibrator’s output. To measure the peak-to-peak voltage of various test stimuli, the sound output from the ABR recording system (through insert earphones) was routed to the sound level meter. A two-channel oscilloscope (Tektronix DPO2012) was connected to the sound level meter to capture the signal. Each stimulus was presented at 20 stimuli/s at 90, 85, and 80 dB nHL. Peak-to-peak voltage for each stimulus from different systems was recorded. Table 2 shows the peak-to-peak voltage for different intensities at different frequencies for different systems.

Table 2.

Peak-to-Peak Voltage (in Millivolts) for Different Test Stimuli at Different Intensities for Different Systems

	Ep-25				IHS				Vivosonic				IHS vs. Vivosonicb	IHS vs. EP-25c	Vivosonic vs. EP-25d
Stimulus	90 dB nHL	85 dB nHL	80 dB nHL	dBa	90 dB nHL	85 dB nHL	80 dB nHL	dBa	90 dB nHL	85 dB nHL	80 dB nHL	dBa
BB-Click	710.3	358.4	202.5	10.90	808.7	453.4	250.3	10.19	785.3	437	245.2	10.11	0.26	1.13	0.87
0.5-kHz TB	195.9	108	63.06	9.85	222	120.5	66.43	10.48	209.4	114	62.87	10.45	0.51	1.09	0.58
1-kHz TB	133.7	72.8	43.41	9.77	155.8	86.2	47.8	10.26	159	89.23	57.04	8.90	−0.18	1.33	1.51
2-kHz TB	284.3	163.7	92.89	9.72	380.7	210	118.2	10.16	293.6	162.6	79.05	11.40	2.26	2.54	0.28
4-kHz TB	365	203.2	114.6	10.06	832.3	450.9	251.1	10.41	324.8	193.8	109.2	9.47	8.17	7.16	−1.01
BB-Chirp	618.7	303.4	158.3	11.84	405.6	233.6	126.8	10.08	246.6	136.5	75.13	10.32	4.30	−3.69	−7.99
0.5-kHz NBC	263.8	146.5	80.38	10.32	185.4	102.7	54.98	10.56	119.2	65.44	45.96	8.28	3.84	−3.06	−6.90
1-kHz NBC	192.8	109.9	60.84	10.02	679.4	379.6	211.1	10.15	118.4	66.54	42.94	8.81	15.18	10.94	−4.24
2-kHz NBC	405.4	230.4	130.7	9.83	1699	939.6	521.3	10.26	290.7	160.2	93.25	9.88	15.33	12.45	−2.89
4-kHz NBC	588.8	329.2	187.2	9.95	1437	798.8	453.3	10.02	310.4	175.5	97.84	10.03	13.31	7.75	−5.56

Note: The dB values are in italics. Columns that show dB values are linearity values. The last three columns show differences in dB values between systems. BB = broadband; NBC = narrowband Chirp; TB = tone burst.

^a 20 × log(peak-to-peak voltage at 90 dB nHL/peak-to-peak voltage at 80 dB nHL).

^b 20 × log(peak-to-peak voltage at 90 dB nHL of the IHS system/peak-to-peak voltage at 90 dB nHL of the Vivosonic system).

^c 20 × log(peak-to-peak voltage at 90 dB nHL of the IHS system/peak-to-peak voltage at 90 dB nHL of the EP-25 system).

^d 20 × log(peak-to-peak voltage at 90 dB nHL of the Vivosonic system/peak-to-peak voltage at 90 dB nHL of the EP-25 system).

Linearity check

For each system, linearity of the stimulus was examined. Differences in peak-to-peak voltage for two different intensity levels for a specific stimulus were calculated and converted to dB (20 × log[peak-to-peak voltage at 90 dB nHL/peak-to-peak voltage at 80 dB nHL]). The dB values are in italics in Table 2. The linearity for all stimuli appears to be satisfactory (±3 dB, IEC 60645-1) for all three systems.

Differences in Sound Output Between Systems

The peak-to-peak voltage for the 90 dB nHL dial setting of a stimulus was compared across systems to determine whether they were consistent. The peak-to-peak voltage of a stimulus from two different systems was converted to dB [20 × log(peak-to-peak voltage at 90 dB nHL of system-1/peak-to-peak voltage at 90 dB nHL of system-2)]. Differences in dB values between systems are reported in the last three columns of Table 2. The dB values for the click, 0.5-, 1-, and 2-kHz tone burst stimuli across all three systems were within ±3 dB. However, the dB values for the 4-kHz tone burst and all the chirp stimuli were ≥3 dB between systems.

Figure 2 shows the frequency response curves of different stimuli for different systems. Frequency response curves are comparable for traditional stimuli, but not for chirp stimuli. Narrowband chirp stimuli of all three systems are broader when compared with traditional tone burst stimuli. The peak-to-peak voltage of the IHS system for iChirp stimuli (for 1-, 2-, and 4-kHz narrowband iChirps) and for 4-kHz tone bursts is higher when compared with the other two manufacturers (refer to Table 2).

Figure 2.

Frequency response curves for different test stimuli for different systems.

Behavioral Thresholds for Different Test Stimuli

Within-System Comparison (Traditional vs. Chirp Stimuli)

Figure 3A shows the behavioral thresholds for traditional and chirp stimuli for different ABR recording systems. A statistical analysis was conducted on comparable stimuli (e.g., broadband click vs. broadband chirp, etc.) to examine any differences in thresholds between traditional and chirp stimuli within a given system. Table 3 shows the statistical results for behavioral thresholds of traditional and chirp stimuli. The integrity V-500 system showed a significantly lower behavioral threshold for broadband click and 4-kHz tone burst stimuli when compared with broadband VF Chirp and 4-kHz narrowband VF Chirp stimuli, respectively. The EP-25 system showed a significantly lower behavioral threshold for broadband CE-Chirp, 0.5-kHz narrowband CE-Chirp, and 1-kHz narrowband CE-Chirp stimuli when compared with broadband click, 0.5-, and 1-kHz tone burst stimuli, respectively. The IHS system showed a significantly lower behavioral threshold for 0.5-, 1-, 2-, and 4-kHz narrowband iChirp stimuli when compared with 0.5-, 1-, 2-, and 4-kHz tone burst stimuli.

Figure 3.

Within-system (A) and between-system (B) comparison of behavioral thresholds for different stimuli. Error bars depict the standard error of the mean.

Table 3.

Statistical Results of the Comparison of Behavioral Thresholds of Traditional and Chirp Stimuli for Each ABR Recording System

EP-25			IHS			Vivosonic
Traditional stimuli (in dB nHL)	Chirp stimuli (in dB nHL)	Statistics	Chirp stimuli (in dB nHL)	Traditional stimuli (in dB nHL)	Statistics	Traditional stimuli (in dB nHL)	Chirp stimuli (in dB nHL)	Statistics
BB-Click (3.66)	BB-Chirp (0.66)	W = 36, z = 2.52, p = 0.008	BB-Click (3.66)	BB-Chirp (4.66)	W = 3, z = −1.21, p = 0.23	BB-Click (0.33)	BB-Chirp (4.33)	W = 0.00, z = −2.66, p = 0.006
0.5 kHz TB (8.33)	0.5 kHz NBC (2.66)	t (14) = 3.52, p = 0.003	0.5 kHz TB (8.66)	0.5 kHz NBC (7)	W = 15, z = 2.02, p = 0.03	0.5 kHz TB (6)	0.5 kHz NBC (4.67)	W = 15, z = 0.94, p = 0.37
1 kHz TB (3.66)	1 kHz NBC (1.33)	W = 28, z = 2.36, p = 0.01	1 kHz TB (7.33)	1 kHz NBC (−5.33)	W = 120, z = 3.40, p < 0.001	1 kHz TB (1.33)	1 kHz NBC (2.67)	W = 3.50, z = −1.46, p = 0.12
2 kHz TB (1.66)	2 kHz NBC (1.00)	W = 13, z = 0.52, p = 0.66	2 kHz TB (5)	2 kHz NBC (−4.33)	W = 105, z = 3.29, p < 0.001	2 kHz TB (1)	2 kHz NBC (2.67)	W = 4.00, z = −1.69, p = 0.07
4 kHz TB (0.66)	4 kHz NBC (0.33)	W = 5.50, z = 0.18, p = 1.00	4 kHz TB (−1.33)	4 kHz NBC (−3.33)	W = 21, z = 2.20, p = 0.01	4 kHz TB (0)	4 kHz NBC (4)	W = 0.00, z = −2.80, p = 0.003

Note: Mean thresholds are reported in dB nHL. BB = broadband; IHS = Intelligent Hearing Systems, EP-25: Eclipse; NBC = narrowband Chirp; TB = tone burst.

Between-System Comparison (Traditional vs. Chirp Stimuli)

Table 4 shows the statistical results of behavioral thresholds compared between different ABR recording systems. Statistical analyses showed significant differences in behavioral thresholds between systems for the broadband click, 1-kHz tone burst stimulus, broadband chirp, and 1-, 2-, and 4-kHz narrowband chirps.

Table 4.

Statistical Results of Behavioral Thresholds Compared Between Different ABR Recording Systems

Stimuli	Statistics	Post hoc comparison with an adjusted p value for each stimulus (0.05/3 = 0.016)
BB-Click	χ2 (2) = 0.16, p = 0.010	IHS vs. EP-25: U = 106.50, p = 0.781 IHS vs. Vivosonic: U = 48.50, p = 0.003 EP-25 vs. Vivosonic: U = 63.50, p = 0.023
0.5-kHz TB	F(2, 42) = 1.45, p = 0.24
1-kHz TB	F(2, 42) = 8.35, p = 0.001	IHS vs. EP-25: t(28) = 2.77, p = 0.010 IHS vs. Vivosonic: t(28) = 3.65, p = 0.001 EP-25 vs. Vivosonic: t(28) = 1.60, p = 0.121
2-kHz TB	F(2, 42) = 2.53, p = 0.09
4-kHz TB	F(2, 42) = 0.58, p = 0.56
BB-Chirp	F(2, 42) = 5.55, p = 0.007	IHS vs. EP-25: t(28) = 2.84, p = 0.008 IHS vs. Vivosonic: t(28) = 0.237, p = 0.815 EP-25 vs. Vivosonic: t(28) = 2.40, p = 0.023
0.5-kHz NBC	F(2, 42) = 2.26, p = 0.116
1-kHz NBC	χ2 (2) = 18.83, p < 0.001	IHS vs. EP-25: U = 36.00, p < 0.001 IHS vs. Vivosonic: U = 26.00, p < 0.001 EP-25 vs. Vivosonic: U = 82.50, p = 0.167
2-kHz NBC	χ2 (2) = 8.80, p = 0.012	IHS vs. EP-25: U = 64.00, p = 0.034 IHS vs. Vivosonic: U = 50.00, p = 0.007 EP-25 vs. Vivosonic: U = 87.50, p = 0.264
4-kHz NBC	F(2, 42) = 11.81, p < 0.001	IHS vs. EP-25: t(28) = 2.36, p = 0.025 IHS vs. Vivosonic: t(28) = 4.55, p < 0.001 EP-25 vs. Vivosonic: t(28) = 2.71, p = 0.011

Note: Statistically significant comparisons are shown in italics. BB = broadband; NBC = narrowband Chirp; TB = tone burst.

ABR Thresholds for Different Test Stimuli

Within-System Comparison (Traditional vs. Chirp Stimuli)

For each ABR recording system, ABR thresholds obtained for traditional and chirp stimuli were compared (e.g., broadband click vs. broadband chirp, etc.). Figure 4 shows the within- and between-system comparison of ABR thresholds for different test stimuli. Table 5 shows the statistical results for ABR thresholds of traditional and chirp stimuli. The integrity V-500 system showed significantly lower ABR thresholds for broadband click and 0.5-kHz tone burst stimuli when compared with broadband VF Chirp and 0.5-kHz narrowband VF Chirp stimuli, respectively. The EP-25 system showed lower ABR thresholds for broadband CE-Chirp and 1-, 2-, and 4-kHz narrowband CE-Chirp stimuli when compared with broadband click and 1-, 2-, and 4-kHz tone burst stimuli. The ABR thresholds obtained from the IHS system were similar for traditional and chirp stimuli.

Figure 4.

Within-system (A) and between-system (B) comparison of ABR thresholds for different stimuli. Error bars depict the standard error of the mean.

Table 5.

Statistical Results of the Comparison of ABR Thresholds of Traditional and Chirp Stimuli for Each ABR Recording System

EP-25			IHS			Vivosonic
Traditional stimuli (in dB nHL)	Chirp stimuli (in dB nHL)	Statistics	Traditional stimuli (in dB nHL)	Chirp stimuli (in dB nHL)	Statistics	Traditional stimuli (in dB nHL)	Chirp stimuli (in dB nHL)	Statistics
BB-Click (14)	BB-Chirp (10.66)	W = 33, z = 2.10, p = 0.03	BB-Click (12.66)	BB-Chirp (12)	W = 37, z = 0.35, p = 0.75	BB-Click (11.6)	BB-Chirp (14.6)	W = 3.50, z = −2.03, p = 0.040
0.5-kHz TB (28)	0.5-kHz NBC (24)	t(14) = 1.18, p = 0.25	0.5-kHz TB (23.6)	0.5-kHz NBC (27.33)	W = 15, z = −1.27, p = 0.21	0.5-kHz TB (27.3)	0.5-kHz NBC (37.66)	t(14) = 5.38, p < 0.001
1-kHz TB (22.66)	1-kHz NBC (15)	t(14) = 3.52, p = 0.003	1-kHz TB (18.66)	1-kHz NBC (13)	W = 55, z = 1.95, p = 0.054	1-kHz TB (18.33)	1-kHz NBC (20.66)	t(14) = 0.89, p = 0.38
2-kHz TB (21.66)	2-kHz NBC (13)	W = 55, z = 2.80, p = 0.005	2-kHz TB (13.33)	2-kHz NBC (8.66)	W = 49, z = 2.19, p = 0.029	2-kHz TB (18)	2-kHz NBC (15)	W = 35, z = 0.76, p = 0.45
4-kHz TB (17.66)	4-kHz NBC (10.33)	t(14) = 3.77, p = 0.005	4-kHz TB (11)	4-kHz NBC (10)	W = 13.5, z = 0.62, p = 0.50	4-kHz TB (17)	4-kHz NBC (14.33)	W = 46.50, z = 1.28, p = 0.19

Note: Mean thresholds are reported in dB nHL. Statistically significant comparisons are shown in italics. BB = broadband; NBC = narrowband Chirp; TB = tone burst.

Between-System Comparison (Traditional vs. Chirp Stimuli)

ABR thresholds obtained for traditional and chirp stimuli from different systems were compared. Table 6 shows the statistical results. There were significant differences in thresholds between ABR recording systems for 2-kHz tone burst and 0.5- and 2-kHz narrowband chirp stimuli.

Table 6.

Statistical Results of ABR Thresholds Compared Between Different ABR Recording Systems

Stimuli	Statistics	Post hoc comparison with an adjusted p value (0.05/3 = 0.016)
BB-Click	F(2, 42) = 0.44, p = 0.64
0.5-kHz TB	F(2, 42) = 0.87, p = 0.42
1-kHz TB	χ2(2) = 2.31, p = 0.31
2 kHz TB	F(2, 42) = 4.21, p = 0.02	IHS vs. EP-25: t(28) = 3.05, p = 0.005 IHS vs. Vivosonic: t(28) = 1.58, p = 0.124 EP-25 vs. Vivosonic: t(28) = 1.24, p = 0.224
4-kHz TB	F(2, 42) = 2.23, p = 0.11
BB-Chirp	F(2, 42) = 1.67, p = 0.19
0.5-kHz NBC	F(2, 42) = 7.19, p = 0.002	IHS vs. EP-25: t(28) = 0.83, p = 0.411 IHS vs. Vivosonic: t(28) = 2.69, p = 0.012 EP-25 vs. Vivosonic: t(28) = 3.00, p < 0.001
1-kHz NBC	F(2, 42) = 2.41, p = 0.102
2-kHz NBC	F(2, 42) = 3.45, p = 0.041	IHS vs. EP-25: t(28) = −1.88, p = 0.07 IHS vs. Vivosonic: t(28) = −2.31, p = 0.02 EP-25 vs. Vivosonic: t(28) = −0.85, p = 0.39
4-kHz NBC	F(2, 42) = 2.13, p = 0.13

Note: Statistically significant comparisons are shown in italics. BB = broadband; NBC = narrowband Chirp; TB = tone burst.

ABR Peak V Latency and Amplitude

Figure 5 shows the peak V latencies for different stimuli and different systems, recorded at 80 dB nHL. The mean latencies are slightly longer for the IHS system for traditional stimuli and significantly longer for chirp stimuli. Table 7 shows the statistical analyses of peak V latency for different stimuli. For traditional stimuli, there were significant differences in peak V latencies for broadband click, 1-, and 2-kHz tone burst stimuli. For chirp stimuli, significant differences in latencies were observed in broadband chirp, 0.5-, 1-, and 2-kHz narrowband chirp stimuli.

Figure 5.

Peak V latencies for traditional and chirp stimuli for different ABR systems, recorded at 80 dB nHL. Error bars depict the standard error of the mean.

Table 7.

Statistical Analysis on Peak V Latencies for Different Test Stimuli

		Peak V Latency in ms (SD)	ANOVA	Post hoc comparison with Bonferroni corrections
Traditional stimuli	BB-Click	EP-25: 5.47 (0.25) Vivosonic: 5.39 (0.24) IHS: 5.66 (0.20)	F(2, 42) = 4.88, p = 0.01	EP-25 vs. Vivosonic: t(42) = 0.18, p = 1.00 EP-25 vs. IHS: t(42) = 2.22, p = 0.10 Vivosonic vs. IHS: t(42) = 3.04, p = 0.01
	0.5 kHz	EP-25: 7.19 (0.72) Vivosonic: 7.19 (0.38) IHS: 7.57 (0.49)	F(2, 32.8) = 2.28, p = 0.11
	1 kHz	EP-25: 6.84 (0.34) Vivosonic: 6.94 (0.29) IHS: 7.18 (0.34)	F(2, 42) = 4.38, p = 0.01	EP-25 vs. Vivosonic: t(42) = 0.85, p = 1.00 EP-25 vs. IHS: t(42) = 2.88, p = 0.01 Vivosonic vs. IHS: t(42) = 2.02, p = 0.14
	2 kHz	EP-25: 6.35 (0.30) Vivosonic: 6.33 (0.33) IHS: 6.61 (0.27)	F(2, 42) = 3.91, p = 0.02	EP-25 vs. Vivosonic: t(42) = 0.10, p = 1.00 EP-25 vs. IHS: t(42) = 2.37, p = 0.06 Vivosonic vs. IHS: t(42) = 2.47, p = 0.052
	4 kHz	EP-25: 6.02 (0.27) Vivosonic: 6.07 (0.30) IHS: 6.13 (0.21)	F(2, 42) = 0.64, p = 0.52
Chirp stimuli	BB-Chirp	EP-25: 5.58 (0.30) Vivosonic: 5.06 (0.50) IHS: 8.23 (0.61)	F(2, 33.6) = 180.71, p < 0.001	EP-25 vs. Vivosonic: t(42) = 2.91, p = 0.01 EP-25 vs. IHS: t(42) = 14.81, p < 0.001 Vivosonic vs. IHS: t(42) = 17.72, p < 0.001
	0.5-kHz NBC	EP-25: 5.36 (0.39) Vivosonic: 6.75 (0.99) IHS: 7.91 (0.48)	F(2, 42) = 54.03, p < 0.001	EP-25 vs. Vivosonic: t(42) = 5.63, p < 0.001 EP-25 vs. IHS: t(42) = 10.38, p < 0.001 Vivosonic vs. IHS: t(42) = 4.74, p < 0.001
	1-kHz NBC	EP-25: 5.71 (0.50) Vivosonic: 6.18 (0.62) IHS: 7.31 (0.40)	F(2, 42) = 38.07, p < 0.001	EP-25 vs. Vivosonic: t(42) = 2.51, p = 0.04 EP-25 vs. IHS: t(42) = 8.49, p < 0.001 Vivosonic vs. IHS: t(42) = 5.97, p < 0.001
	2-kHz NBC	EP-25: 5.56 (0.40) Vivosonic: 5.90 (0.46) IHS: 6.45 (0.27)	F(2, 42) = 20.21, p < 0.001	EP-25 vs. Vivosonic: t(42) = 2.35, p = 0.06 EP-25 vs. IHS: t(42) = 6.29, p < 0.001 Vivosonic vs. IHS: t(42) = 3.93, p < 0.001
	4-kHz NBC	EP-25: 5.58 (0.35) Vivosonic: 5.60 (0.38) IHS: 6.02 (0.79)	F(2, 42) = 3.11, p = 0.052

Note: Statistically significant comparisons are shown in italics. ANOVA = analysis of variance; BB = broadband; NBC = narrowband Chirp.

Figure 6 shows the peak V amplitudes for different stimuli and different systems, recorded at 80 dB nHL. Table 8 shows the statistical analyses for the peak V amplitudes. For traditional stimuli, peak V amplitude levels were significantly different between systems for 0.5-, 1-, and 2-kHz tone burst stimuli. For chirp stimuli, peak V amplitude differences between systems were observed for broadband chirp, 1-, and 2-kHz narrowband chirp stimuli. Figure 7 shows the grand average ABRs for different stimuli recorded from three different systems.

Figure 6.

Peak V amplitudes for traditional and chirp stimuli for different ABR systems, recorded at 80 dB nHL. Error bars depict the standard error of the mean.

Table 8.

Statistical Analysis of Peak V Amplitude for Different Test Stimuli

		Peak V in microvolts (SD)	ANOVA	Post hoc comparison with Bonferroni corrections
Traditional stimuli	BB-Click	EP-25: 0.75 (0.27) Vivosonic: 0.57 (0.21) IHS: 0.67 (0.18)	F(2, 42) = 2.19, p = 0.12
	0.5 kHz	EP-25: 0.65 (0.25) Vivosonic: 0.38 (0.14) IHS: 0.65 (0.25)	F(2, 36.3) = 7.10, p = 0.002	EP-25 vs. Vivosonic: t(42) = 3.18, p = 0.008 EP-25 vs. IHS: t(42) = 0.14, p = 0.01 Vivosonic vs. IHS: t(42) = 3.33, p = 0.005
	1 kHz	EP-25: 0.54 (0.21) Vivosonic: 0.41 (0.15) IHS: 0.64 (0.21)	F(2, 42) = 4.92, p = 0.01	EP-25 vs. Vivosonic: t(42) = 1.74, p = 0.26 EP-25 vs. IHS: t(42) = 1.38, p = 0.51 Vivosonic vs. IHS: t(42) = 3.13, p = 0.009
	2 kHz	EP-25: 0.55 (0.14) Vivosonic: 0.41 (0.14) IHS: 0.56 (0.14)	F(2, 42) = 5.17, p = 0.009	EP-25 vs. Vivosonic: t(42) = 2.71, p = 0.02 EP-25 vs. IHS: t(42) = 0.12, p = 1.00 Vivosonic vs. IHS: t(42) = 2.84, p = 0.02
	4 kHz	EP-25: 0.46 (0.18) Vivosonic: 0.38 (0.17) IHS: 0.45 (0.21)	F(2, 42) = 0.71, p = 0.49
Chirp stimuli	BB-Chirp	EP-25: 0.81 (0.31) Vivosonic: 0.47 (0.21) IHS: 0.75 (0.36)	F(2, 42) = 5.57, p = 0.007	EP-25 vs. Vivosonic: t(42) = 3.14, p = 0.009 EP-25 vs. IHS: t(42) = 0.58, p = 1.00 Vivosonic vs. IHS: t(42) = 2.55, p = 0.04
	0.5-kHz NBC	EP-25: 0.38 (0.23) Vivosonic: 0.39 (0.10) IHS: 0.55 (0.28)	F(2, 42) = 2.91, p = 0.06
	1-kHz NBC	EP-25: 0.63 (0.33) Vivosonic: 0.37 (0.10) IHS: 0.64 (0.30)	F(2, 30.61) = 4.81, p = 0.01	EP-25 vs. Vivosonic: t(42) = 2.61, p = 0.03 EP-25 vs. IHS: t(42) = 0.14, p = 1.00 Vivosonic vs. IHS: t(42) = 2.75, p = 0.02
	2-kHz NBC	EP-25: 0.64 (0.17) Vivosonic: 0.42 (0.18) IHS: 0.56 (0.16)	F(2, 42) = 6.13, p = 0.004	EP-25 vs. Vivosonic: t(42) = 3.45, p = 0.003 EP-25 vs. IHS: t(42) = 1.22, p = 0.67 Integrity V500 vs. IHS: t(42) = 2.22, p = 0.09
	4-kHz NBC	EP-25: 0.52 (0.21) Vivosonic: 0.40 (0.14) IHS: 0.48 (0.16)	F(2, 42) = 1.99, p = 0.14

Note: Statistically significant comparisons are shown in italics. ANOVA = analysis of variance; BB = broadband; NBC = narrowband Chirp.

Figure 7.

Grand average ABRs for traditional and chirp stimuli at 80 dB nHL recorded from different ABR systems. The IHS does not time-adjust the ABRs for chirp stimuli, whereas the EP-25 and Vivosonic systems do time-adjust the ABRs.

DISCUSSION

In this study, three FDA-approved ABR recording systems were compared. Using these three systems, behavioral and ABR thresholds and peak V latencies and amplitudes (at 80 dB nHL) for traditional and chirp stimuli were examined.

The peak-to-peak voltage of different stimuli at different intensities of all three systems was examined. The linearity of different stimuli for different systems was assessed, and it was within the tolerance limits for all three systems, indicating that these systems are operating predictably. However, there were significant differences in the output between systems, especially for chirp stimuli. The IHS system showed a larger peak-to-peak amplitude for chirp (1-, 2-, and 4-kHz narrowband iChirp) and 4-kHz tone burst stimuli compared with other systems. We hypothesize that these differences could be because of differences in the dBpeSPL values (corresponding to dB nHL) set for these stimuli in the IHS system. These level differences may account for lower behavioral thresholds with the IHS system for these respective stimuli. However, there was no significant difference in ABR thresholds for these stimuli when compared with other systems, although there was a trend for lower ABR thresholds at these test frequencies. Also, the peak-to-peak voltage differences between traditional and chirp stimuli at 90 dB nHL were converted to dB SPL and compared within each ABR system. Most of the EP-25 and the IHS system’s chirp stimuli have higher SPLs than traditional stimuli, whereas the Vivosonic Integrity V500 system’s traditional stimuli have higher SPLs compared with chirp stimuli. Observed SPL differences between traditional and chirp stimuli may result in variations in both behavioral and ABR thresholds.

Behavioral Threshold Differences

The EP-25 system showed a lower behavioral threshold for broadband CE-Chirp and 0.5- and 1-kHz narrowband CE-Chirp stimuli when compared with broadband click, 0.5-, and 1-kHz tone burst stimuli. The IHS system showed a lower threshold for 0.5-, 1-, 2-, and 4-kHz narrowband iChirp stimuli when compared with 0.5-, 1-, 2-, and 4-kHz tone burst stimuli. Conversely, the Integrity V500 system showed a lower behavioral threshold of broadband click and 4-kHz tone burst when compared with broadband chirp and 4-kHz narrowband chirp stimuli. ABR systems that showed differences in behavioral thresholds between traditional and chirp stimuli were within ∼5 to 10 dB.

System differences in behavioral thresholds were observed for test stimuli between systems. The overall threshold differences between systems were on the order of 5 to 10 dB. Behavioral thresholds obtained from the Integrity V500 system for broadband clicks were significantly lower when compared with thresholds obtained from the IHS system. The behavioral thresholds obtained from the IHS system for 1-kHz tone burst stimuli were significantly higher when compared with thresholds obtained from the EP-25 and Integrity V-500 systems. Behavioral thresholds of the IHS broadband iChirp were significantly lower when compared with broadband CE-Chirp thresholds. The behavioral thresholds of the IHS system for 1-, 2-, and 4-kHz narrowband iChirp were significantly lower when compared with thresholds of narrowband chirp stimuli of EP-25 (CE-Chirp) and Integrity V500 (VF Chirp) systems. It is not surprising to see differences in behavioral thresholds for chirp stimuli between systems because of differences in peak-to-peak voltage between these stimuli.

Behavioral thresholds obtained from this study cannot be directly compared with behavioral thresholds reported by the manufacturer or with other previously published studies (Fedtke & Richter, 2007; Gøtsche-Rasmussen et al., 2012) because of differences in the stimulus rates and step size used for behavioral threshold estimation.

ABR Threshold Differences

ABR thresholds for traditional and chirp stimuli were compared within a given ABR system. Overall, the EP-25 system’s ABR thresholds were significantly lower for CE-Chirp stimuli when compared with traditional stimuli, except for the 0.5-kHz test frequency. Previous studies that have used the EP-25 system reported similar findings (Ferm et al., 2013; Ferm & Lightfoot, 2015). However, in our study, ABR thresholds for 0.5-kHz narrowband CE-Chirp were lower, but not statistically significant, when compared with 0.5-kHz tone burst stimuli for the EP-25 system. The Integrity V-500 showed lower ABR thresholds for broadband click and 0.5-kHz tone burst stimuli when compared with broadband VF Chirp and 0.5-kHz narrowband VF Chirp stimuli, respectively. At other test frequencies, there were no significant differences in ABR thresholds. The ABR thresholds obtained using the traditional and chirp stimuli were similar in the IHS system, except for the 2-kHz test stimuli.

Between-system comparison of ABR thresholds revealed significant differences in ABR thresholds at 2-kHz tone burst and 0.5- and 2-kHz narrowband chirp stimuli. For a 2-kHz tone burst stimulus, ABR thresholds obtained from the IHS system were significantly lower when compared with thresholds obtained from the EP-25 system. For 0.5-kHz narrowband chirp stimuli, the Integrity V-500 system showed higher ABR thresholds compared with EP-25 and IHS systems. For a 2-kHz narrowband chirp, ABR thresholds obtained from the IHS system were lower when compared with the other two systems; however, post hoc comparisons were not significant.

The ABR threshold results for traditional test stimuli in the Integrity V-500 system for this study align with those of a recently published system comparison study (Bagatto et al., 2024). Bagatto et al. (2024) reported no significant differences in ABR thresholds for traditional frequency-specific tone bursts between the Integrity V500 and the Biologic Navigator Pro systems. The ABR threshold results for chirp stimuli in the EP-25 and the IHS systems from this study are inconsistent with a recent comparison study of similar recording systems (Leusin Mattiazzi et al., 2024). The ABR thresholds obtained for 2-kHz narrowband iChirp (IHS system) are lower when compared with thresholds obtained for LS 2-kHz narrowband CE-Chirp (EP-25 system) stimuli. However, Leusin Mattiazzi et al. (2024) reported no significant differences in ABR thresholds between narrowband (0.5-, 1-, 2-, and 4-kHz) iChirp and CE-Chirp stimuli. Additional studies are required to resolve these discrepancies.

The findings of this study suggest that ABR thresholds can vary between ABR systems, for both traditional and chirp stimuli. The ABR threshold differences between systems can be ±5 to 10 dB nHL, especially for some narrowband chirp stimuli. We hypothesize that the observed differences in behavioral and ABR thresholds may arise from two factors: the different methods used by each system to generate chirp stimuli, and the variations in dBpeSPL values (corresponding to dB nHL) of these stimuli between systems. The duration of CE-Chirp used in the EP-25 system varies depending on the intensity level, whereas the iChirp (IHS system) and VF-Chirp duration do not change as a function of intensity. Ideally, the peak-to-peak voltage of stimuli used to estimate thresholds should be similar between systems. If peak-to-peak voltage differences between different systems are not within tolerance limits (±3 dB), it can lead to variation in thresholds, which was observed in this study. In other words, the ABR thresholds may vary depending on the system being used, and audiologists should be aware of these differences.

The ABR test is used to estimate behavioral hearing thresholds in infants. In infants with hearing loss, these behavioral thresholds are used to prescribe amplification (Bagatto et al., 2010). The ABR thresholds estimated from one system should be comparable with thresholds obtained from other systems. Inappropriate diagnosis of hearing loss can be made if there are clinically significant differences in ABR thresholds between two different ABR systems. If behavioral thresholds estimated from the ABR thresholds are inaccurate, it can lead to underamplification or overamplification, which can negatively impact speech and language development in children with hearing loss (Walker, 2023). However, the clinical relevance of this difference in threshold between ABR systems is currently not well known and merits further research.

Currently, there is no ANSI standard for calibrating brief-duration stimuli that are typically used in threshold assessments. Therefore, it is feasible that slight differences may occur among different clinical instruments because of differences in calibration. Thus, the need exists to develop equipment-specific data across various stimuli to determine typical minimum response levels behaviorally and electrophysiologically (Atcherson & Stoody, 2012; Hall, 2007). However, some researchers argue with this method and propose standardized reference thresholds (Stapells, 2015). Alternatively, one approach to control these variations in thresholds has been adopted by some infant hearing programs, i.e., using one manufacturer’s ABR system and a standardized protocol for estimating hearing thresholds (British Columbia Early Infant Hearing Program [Hatton et al., 2022]; Ontario Infant Hearing Program [Bagatto et al., 2010]). These programs receive government funding, however, and this would be difficult to implement in countries that do not receive this funding from their respective governments. Furthermore, manufacturer-specific corrections and protocols also can be used.

Currently, the preferred stimuli (traditional vs. chirp stimuli) used by audiologists to estimate hearing thresholds are not universally known. The audiologist must be aware of these differences between systems when interpreting data provided by other facilities using equipment different from their own.

Using the data from this study, we calculated the eHLs (ABR thresholds in dB nHL − the pure-tone behavioral threshold in dB HL) and compared them with published studies (refer to Table 9). The mean eHLs for tone burst stimuli for all systems of this study are more closely comparable with published data (Stapells, 2000). The mean eHLs for narrowband CE-Chirp stimuli obtained with the EP-25 system are similar to that of the Sininger et al. (2020) study that also used a similar system. We could not find published data for these stimuli for the IHS and the Integrity V-500 systems.

Table 9.

eHLs or the Correction Factors from This Study Are Compared with Previously Published Studies

Traditional Tone Burst Stimuli (dB)					Narrowband Chirp Stimuli (dB)
Study	0.5 kHz	1 kHz	2 kHz	4 kHz	Study	0.5 kHz	1 kHz	2 kHz	4 kHz
Stapells (2000)	−20	−16	−13	−12	Sininger et al. (2020) EP-25	−15	−10	−5	−5
Vivosonic (current study)	−21	−11	−16	−13	Vivosonic (current study)	−31	−14	−12	−11
EP-25 (current study)	−21	−17	−18	−14	EP-25 (current study)	−18	−9	−9	−6
IHS (current study)	−17	−12	−10	−7	IHS (current study)	−21	−7	−5	−6

Several studies have been conducted to examine CE-Chirp evoked ABRs in different clinical populations (Bargena, 2015; Ferm et al., 2013; Ferm & Lightfoot, 2015; Sininger et al., 2018, 2020). Correction values are available for different transducers and different age ranges, and correction values are adjusted based on the severity of hearing loss (Sininger et al., 2018, 2020). However, there is a scarcity of studies examining iChirp and VF Chirp evoked ABRs in individuals with different degrees of hearing loss and different age ranges.

ABR Latency and Amplitude

Significant differences in peak V latencies between systems were observed. Differences were observed for stimuli that are used for neurodiagnostic (broadband clicks and chirp) and threshold estimation (frequency-specific stimuli) purposes. Peak V latencies were significantly longer in the IHS system for the broadband click and 1- and 2-kHz tone burst, as well as for broadband chirp and narrowband (0.5-, 1-, 2-, and 4-kHz) chirp stimuli when compared with the other two systems. Peak V latencies of the IHS system for traditional stimuli were between 0.20 and 0.40 millisecond (ms) longer when compared with the other two systems. We hypothesize that this could be because of differences in sampling points between ABR systems. The ABRs recorded in the IHS system have higher sampling points (n = ∼1,024) when compared with the other two systems (EP-25: n = ∼467 sampling points; Integrity V500: n = ∼240 sampling points). Whereas Peak V latencies of the IHS system for chirp stimuli were between 0.40 and 3.17 ms longer when compared with the other two systems. This could be because of a combination of differences in the number of sampling points and the fact that the IHS does not time-adjust the ABRs for chirp stimuli, whereas the other two systems do. All three systems use a set number of sampling points regardless of the ABR recording window. It is important to consider the sampling points and recording window, because these two combinations can affect latency values. Accuracy in latency values is important, especially in neurodiagnostic assessments.

The mean peak V latencies recorded using ABR systems with different sampling rates may fall within 2 standard deviations (SDs) of each other. However, the 2 SD range for the peak V latencies varies depending on the number of sample points used to record the ABRs. Different SD ranges can affect clinical decision-making, particularly when determining whether a person’s peak V latency falls within or outside the normative range. The authors have outlined clear guidelines for establishing normative data for ABRs (Hood, 1998; Sininger, 1992). In addition to these recommendations, it is our opinion that clinicians should be cautious about the number of sampling points used in ABR recordings, because this can influence ABR latencies. Therefore, the number of sampling points and the recording window must be similar between the studies referenced by the authors and the system being used.

For neurodiagnostic purposes, click-evoked ABRs are recorded at higher intensity levels and different rates (Hall, 2007; Hood, 1998). A significant emphasis is placed on the latency and amplitude of ABR peaks (Allen & Allan, 2014; Ankmnal-Veeranna et al., 2019; Hall, 2007; Hood, 1998; Veeranna et al., 2021, 2022). The ABR latencies should occur within a designated time period. If there is any delay in the ear or the auditory brainstem structures, it can affect the timing of the ABR peaks (Hall, 2007; Hood, 1998). If ABR peak latencies are not within 2 SD, it is considered clinically longer (Allen & Allan, 2014; Ankmnal-Veeranna et al., 2019; Hood, 1998). Some researchers recommend the use of published normative data (Hall, 2007; Stapells, 2011). It is appropriate to refer to published norms if the tester’s ABR system is the same as the one used in the study that published normative data, and if the age range, gender of the participants, and ABR recording protocol are the same. However, if the ABR system differs, then the audiologists should be cautious.

For threshold estimation, there is less emphasis on absolute latency, because it is primarily used to estimate the point at which the peak V may occur. The peak V latency value is not compared with normative data. However, changes in peak V latency and amplitude assist clinicians in differentiating neural responses from background EEG noise. Similarly, the peak V latency depends on the frequency of the stimuli; latency is longer for 0.5-kHz when compared with 4-kHz tone-burst stimuli. The clinician may use this knowledge in identifying the peak V latency. Published norms for peak V latencies are not that common for frequency-specific ABRs. However, some systems (e.g., EP-25 and Integrity V500) have incorporated features within the ABR recording software to assist clinicians in detecting peak V for different test stimuli.

For threshold estimation purposes, the latency differences between systems for traditional stimuli may not garner much attention because of small variations in latencies. However, a clinician who is used to recording ABRs for traditional stimuli and switches to chirp stimuli may notice longer latencies for chirp stimuli in the IHS system, potentially leading to a misdiagnosis. Hence the clinician should be aware of these differences between systems.

The ABR amplitude represents the number of neural elements that are activated synchronously. Any changes in the number of neural elements and synchronicity can affect the amplitude (Hall, 2007; Hood, 1998). The peak-to-peak amplitude is highly variable among participants. Hence for neurodiagnostic purposes, the V/I amplitude ratio is recommended (Hall, 2007; Hood, 1998). The ABR amplitude can be affected by the signal-averaging technique used in the ABR system. The signal-averaging technique is mainly used to differentiate electrophysiological responses from the background EEG noise. The three systems used in this study used three different types of signal-averaging techniques. Hence observed amplitude differences between systems could be caused by differences in the signal-averaging techniques.

This study has some limitations. The authors were not blinded to the behavioral and ABR thresholds, which may have introduced bias. The findings of the study primarily apply to individuals whose hearing thresholds are ≤25 dB HL. Further research is required to compare hearing thresholds obtained from different systems in individuals whose hearing thresholds are ≥25 dB HL. In addition, the findings may be limited to studies using a similar recording protocol to the one employed in this research. Although the findings of this study suggest system-dependent differences in threshold, peak V latencies, and amplitude, we also acknowledge that factors such as sampling variability and test-retest reliability may have contributed to the observed differences.

Abbreviations

ABR

auditory brainstem response

ANOVA

analysis of variance

ANSI

American National Standards Institute

CE-Chirp®

Claus Elberling Chirp®

dB

decibel

dB HL

decibels hearing level

dBpeSPL

decibels peak equivalent sound pressure level

EEG

electroencephalography

eHL

estimated hearing level

EP-25

Eclipse-25

FDA

US Food and Drug Administration

h

hour

HL

hearing level

IEC

International Electrotechnical Commission

IHS

Intelligent Hearing System

kHz

kilohertz

LS

level-specific

ms

millisecond

peSPL

peak equivalent sound pressure level

s

second

SD

standard deviation

SPL

sound pressure level